← 返回 Skills 市场
Tool:
237
总下载
0
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install agent-workflow
功能描述
A structured workflow plugin for OpenClaw agents. Guides work through brainstorm → plan → execute → verify → deliver with persistent state, branching, parall...
使用说明 (SKILL.md)
Agent Workflow
A structured workflow engine for OpenClaw agents. Migrated and generalized from the superpowers workflow system into a code-agnostic, general-purpose workflow plugin.
What it does
Provides a persistent state machine that guides your agent through a complete work lifecycle:
brainstorming → writing-plans → [execute] → verification → finishing-work
↓
subagent-driven-execution
OR
executing-plans
With support for:
- Persistent state — workflow survives session restarts
- Multi-project — run multiple workflows concurrently
- Branching — choose execution strategy at branch points
- Context-plugins — fork into review/parallel-agents without leaving main flow
- Soft-guard goto — jump to any step with warnings about skipped prerequisites
- 11 bundled Skills — covering the full workflow lifecycle
Installation
This is a Plugin, not a plain Skill. Install via:
openclaw plugins install clawhub:agent-workflow
openclaw gateway restart
Then enable in your ~/.openclaw/openclaw.json:
{
"plugins": {
"allow": ["agent-workflow"]
},
"tools": {
"allow": ["agent_workflow"]
}
}
Usage
In your agent (via Feishu, Discord, or any channel):
Start a new workflow for my Q2 planning project
The agent will call agent_workflow with action: "start" and guide you through the workflow.
Tool: agent_workflow
| Action | Description |
|---|---|
start |
Begin a new workflow |
status |
View current state (all active workflows if no ID given) |
next |
Advance to the next step |
goto |
Jump to any node (soft-guard warns about skipped steps) |
complete |
Mark current node done |
fork |
Activate a context-plugin without leaving main flow |
join |
Complete a fork and return |
getSkill |
Load full SKILL.md for the current node |
list |
List all workflows |
abandon |
Abandon a workflow |
Bundled Skills
brainstorming— Turn ideas into specswriting-plans— Break specs into tasksexecuting-plans— Sequential executionsubagent-driven-execution— Parallel subagent executionverification-before-completion— Evidence before claimsfinishing-work— Delivery optionsdispatching-parallel-agents— Fork independent tasksrequesting-review— Dispatch reviewer subagentreceiving-review— Evaluate feedback rigorouslysystematic-problem-solving— Root-cause diagnosiswriting-skills— Create/improve Skills
安全使用建议
This plugin appears to be what it claims — a persistent workflow engine that reads project files, writes workflow state, and dispatches subagents. Before installing: 1) Verify the plugin source (clawhub:agent-workflow) is trusted; if unsure, inspect the full source (especially src/state-store.ts and src/skill-loader.ts) for any network calls or unexpected file access. 2) Check the package contents that will be installed (the package-lock shows many SDKs not in package.json) to ensure no surprising third-party binaries are included. 3) Confirm you are comfortable with workflow state being written under ~/.openclaw (or set storageDir to a controlled location). 4) If you do not want autonomous agent invocation, keep disable-model-invocation or other policy controls in mind at the platform level. If you want, provide the contents of src/state-store.ts and src/workflow-engine.ts for a deeper review of file I/O and any network behavior.
功能分析
Type: OpenClaw Skill
Name: agent-workflow
Version: 1.0.0
The bundle is a comprehensive workflow management plugin for OpenClaw agents, providing a persistent state machine to guide complex tasks. The TypeScript code (index.ts, src/state-store.ts) implements standard filesystem-based state management and tool registration without any suspicious network calls or shell executions. The bundled SKILL.md files use authoritative language to enforce process discipline (e.g., requiring root cause analysis and verification), but these instructions are strictly aligned with the stated goal of improving agent reliability and do not contain malicious prompt injections or instructions to exfiltrate data.
能力评估
Purpose & Capability
The name/description (Agent Workflow) matches the code and SKILL.md: it registers an agent_workflow tool, provides workflow steps, and bundles SKILLs for brainstorming → plan → execute → verify → deliver. The plugin stores state (persistent workflows) and loads bundled skills — these capabilities are expected for a workflow manager. There are no declared env vars or external credentials that don't belong to this purpose.
Instruction Scope
SKILL.md explicitly tells agents to "explore project context — check existing files, docs, prior work" and to save artifacts (e.g., docs/specs/...). That file- and state-oriented behavior is coherent for a workflow plugin, but it does mean the plugin/skills will read and write workspace files and may dispatch subagents. Verify that behavior is acceptable for your environment (e.g., private repos, secrets in project files).
Install Mechanism
The SKILL.md instructs installation via the OpenClaw plugin hub (openclaw plugins install clawhub:agent-workflow). The repository provides code files rather than requiring arbitrary remote downloads or unusual install scripts. No extract-from-URL downloads or shorteners were present in the provided metadata.
Credentials
No environment variables, credentials, or config paths are required by the skill. One minor oddity: the included package-lock.json lists many SDKs (Anthropic, AWS clients, etc.) that are not present in package.json dependencies; these entries appear as dev or incidental entries in the lockfile. This is not necessarily malicious, but you should confirm the final published package contents and dev/vendor dependencies before installing from an untrusted source.
Persistence & Privilege
The plugin stores state under a storageDir (defaults to ~/.openclaw/workspace) and is not marked always:true. It requests normal plugin permissions (enable in openclaw.json and allow the agent_workflow tool). This level of persistence is expected for a workflow engine; ensure the default storage location and permissions are acceptable for your environment.
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install agent-workflow - 安装完成后，直接呼叫该 Skill 的名称或使用
/agent-workflow触发 - 根据 Skill 的参数说明提供必要输入，即可获得结构化输出
版本历史
v1.0.0
Agent Workflow v1.0.0 — Initial release. A structured workflow plugin for OpenClaw agents. Migrated and adapted from the superpowers workflow system (originally designed for Claude Code) into a general-purpose, code-agnostic workflow engine. Features: full workflow state machine (brainstorm → plan → execute → verify → deliver), persistent state across sessions, multi-project concurrency, branch support, context-plugins for parallel forks, soft-guard goto, and 11 bundled Skills covering the full workflow lifecycle. Includes agent_workflow tool with actions: start, status, next, goto, complete, fork, join, getSkill, list, abandon.
元数据
常见问题
Agent Workflow 是什么？
A structured workflow plugin for OpenClaw agents. Guides work through brainstorm → plan → execute → verify → deliver with persistent state, branching, parall... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 237 次。
如何安装 Agent Workflow？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install agent-workflow」即可一键安装，无需额外配置。
Agent Workflow 是免费的吗？
是的，Agent Workflow 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
Agent Workflow 支持哪些平台？
Agent Workflow 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（cross-platform）。
谁开发了 Agent Workflow？
由 nano（@kangyishuai）开发并维护，当前版本 v1.0.0。
推荐 Skills